Chapter 9Tax Expenditure Reporting and DomesticRevenue Mobilization in Africa

Agustin Redonda, Christian von Haldenwang, and Flurim Aliu

9.1 Tax Expenditure and Tax Expenditure Reporting

Governments pursue public policy objectives through revenue collection and publicspending. However, they also rely on tax expenditures (TEs). The term refers tobenefits granted to specific sectors or groups through preferential tax treatments suchas exemptions, deductions, credits, deferrals and lower tax rates. Governments useTEs to promote economic growth and attract investments but also for the pursuit ofsocial welfare objectives and to incentivise specific patterns of behaviour, such asenergy consumption from renewable sources.

TEs are hence used widely as public policy instruments, and they are costly. Inthe United States, the federal government is estimated to forego more than USD1.5trillion in 2019, an amount equal to 44% of tax revenue collection and slightly morethan 7% of gross domestic product (GDP) (US Treasury 2020). Existing estimates,though limited in scope, show that TEs range from 0.7 to 6.6% of GDP in LatinAmerica and from 3.3 to 7.5% of GDP in Africa (World Bank 2015).

Yet, despite their significant fiscal cost, TEs are often very opaque. Transparencyon the range and magnitude of existing TEs is limited, and new provisions are beingintroduced regularly without adequate scrutiny. This lack of scrutiny may sometimesbe justified by the urgency of the situation. For instance, all over the world govern-ments have introduced tax reliefs as a short-term response to the coronavirus disease

A. Redonda (*) · F. AliuCouncil on Economic Policies, Zürich, Switzerlande-mail: ar@cepweb.org

C. von HaldenwangGerman Development Institute, Deutsches Institut für Entwicklungspolitik (DIE), Bonn,Germanye-mail: christian.vonhaldenwang@die-gdi.de

© The Author(s) 2021I. J. Mosquera Valderrama et al. (eds.), Taxation, International Cooperation and the2030 Sustainable Development Agenda, United Nations University Series onRegionalism 19, https://doi.org/10.1007/978-3-030-64857-2_9

173

2019 (COVID-19) pandemic.1 However, in a majority of cases, TEs are used topursue more strategic goals, and they tend to persist over time. This makes the issueof TE reporting and transparency a very relevant issue, particularly for low-incomecountries with limited capacities for domestic revenue mobilization (DRM).

Official reports on TEs are based on very heterogeneous standards both withrespect to the quality and the scope of the data they provide. In France, for instance,the budget appendix dedicated to TEs explicitly acknowledges that 254 TEs out of474 cannot be quantified, or only a rough order of magnitude can be given (Directiondu Budget 2019). Switzerland published its last report on federal TEs in 2011(Département Fédéral des Finances 2011). Australia, despite being at the forefrontin terms of TE reporting, does not provide estimates for 49% of all TEs (143/289) inits Tax Expenditures Statement 2017 (Australian Treasury 2018). The picture issignificantly worse when it comes to developing economies. In regions such as theMiddle East and North Africa, less than 10% of the countries report on TEs on aregular basis (World Bank 2015).

As a result, TEs are hardly ever subject to sound cost-benefit analyses. Indeed,empirical studies assessing the effectiveness and efficiency of TEs are rare, and,when available, results are often not the expected ones.2 At the same time, identi-fying and phasing out costly and ineffective provisions are crucial in order toimprove the equity and fairness of the tax system, to increase the mobilization ofdomestic resources and to raise the quality of public expenditure systems. TE reformhas become even more crucial in the current situation, when governments will haveto cope with the medium- and long-term effects of the COVID-19 crisis.

In South Africa, for instance, the government grants pension-related TEs topromote the growth of pension savings and to provide the elderly with a stableincome upon retirement. According to the latest TE report, in 2016 these provisionswere the largest TE in South Africa, amounting to 72,991 million Rand (ca. 5.1billion euros), i.e. 35% of total TEs. Pension-related TEs have been at the heart of anintense debate. The South African government has been trying to deal with a policydesign trade-off between providing generous benefits to boost individual savings andthe risk of compromising the equity of the tax system due to the excessive use ofthese provisions, particularly by those at the top end of the income distribution. Asacknowledged in a Technical Discussion Paper published by the South AfricanTreasury in 2012, “the barriers to a more effective tax incentive regime are the

1An overview over tax measures taken to confront the pandemic can be found at https://www.oecd.org/tax/tax-policy/, accessed on 27.05.2020.2For instance, many tax incentives implemented to attract investment have been proven to performpoorly. This is the case of several tax incentives granted by low-income economies such as taxholidays and special economic zones (WB 2017) as well as patent boxes (Alstadsæter et al. 2018;Klemens 2017) and other tax incentives to boost R&D. Yet, the lack of effectiveness of TEs goesbeyond tax benefits for businesses. The mortgage interest deduction (Hilber and Turner 2014) andthe myriad of tax benefits to boost pension savings (Duflo et al. 2006) have also been proven to beineffective. For a more detailed overview on the lack of effectiveness and efficiency of TEs, seeRedonda (2016).

174 A. Redonda et al.

complexity of the current regime [. . .] as well as the fact that the regime is open toabuse through excessive contributions by employers and high-income earningindividuals. [. . .] The tax exemption has no nominal monetary cap in the case ofhigher-income employees, allowing them to make tax-exempt contributions way inexcess of the amount required to maintain a reasonable standard of living inretirement” (South African National Treasury 2012). A comprehensive reform wasimplemented in 2016 to simplify and harmonise the pension system in the hope thatthis would improve the fairness of the whole system. Yet, as shown by Axelson et al.(2020), pension-related TEs are still highly regressive in the country.

The goal of this chapter is twofold: first, to discuss the role of TE reporting as akey component of a more effective and fairer tax system as well as a determinant forDRM and, second, to provide an overview of TE reporting in Africa.

9.2 The Fiscal Cost and (in)Effectiveness of TaxExpenditures

Revenue mobilisation is vital for many low- and middle-income economies. Thefinancing gap for achieving the Sustainable Development Goals (SDGs) for devel-oping countries alone has been estimated at roughly USD2.5 trillion (UNCTAD2014). External financing—including official development assistance and foreigndirect investment (FDI)—is certainly an important source of revenue. Yet, themobilisation of domestic resources should be considered the most important revenuesource to support inclusive and sustained economic growth (Brys et al. 2016). Taxrevenues amounted to USD4.3 trillion in 2016 for low- and middle-income countriesalone, which is more than double the amount of international public and privatecapital these countries received in the same year (von Haldenwang and Laudage2019).

Indeed, the mobilisation of domestic resources through taxation is particularlyrelevant for many reasons. Revenues collected through the tax system are generallymore stable and predictable than those coming from foreign aid or domestic non-taxsources, e.g. royalties from the minerals sector. Moreover, the contribution of taxescan strengthen the social contract between citizens and their government and thushave a positive effect on governance. Unfortunately, though, tax-to-GDP ratios inmany low- and middle-income economies are strikingly low. Whereas the averagetax-to-GDP ratio for advanced economies is 26%, about half of developing countriescollect less than 15% of GDP in taxes—a threshold that has recently been identifiedas the minimum required to allow developing economies to take off economically(IMF 2017a).

Increasing tax revenues is an ambitious and complex task. It entails working onissues such as designing efficient tax systems, improving institutional capacity,increasing tax compliance, cooperating with other revenue bodies at an internationalscale and empowering tax administrations at various government levels, among

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 175

others. Beyond these technical issues, however, fiscal contract and fiscal sociologydebates show that improving tax systems is above all a systemic task that involvesaddressing the quality of public policies, the strengthening of governmental checksand balances and the legitimacy of government in general (Bird and Zolt 2015;Prichard 2019).

It is against this background that the frequent and sustained use of TEs should beanalysed. While ideological positions may play a role, for instance, with regard tothe private or public provision of social security or health insurance, it can beassumed that in most developing countries, TEs are employed for two main reasons:tax competition and political convenience. Tax competition refers to the fact thatcapital-hungry countries are under pressure to grant incentives for investments to thedegree that this is an established practice among its competitors. Convenience meansthat these instruments are readily available to reward specific clientele groups or toobtain rents. This makes them an important pillar of the politics of taxation and thepolitical economy of tax bargaining.

Phasing out ineffective TEs—which appears a low-hanging fruit for developingcountries at first glance—is in fact a difficult, at times even risky undertaking.Institutional and power lock-ins evolve around specific TEs, making it politicallycostly to dismantle them. Public protests are widely documented in the literature,particularly with regard to fossil fuel subsidies, which often have a large impact onthe costs of living of poorer urban households (for instance, see Chelminski 2018).Protests are often fuelled by a general lack of trust in government, as citizens refuseto trade present benefits for a promise of uncertain future compensations. In addition,powerful economic groups or large individual companies lobby for the extension ofsector- or firm-specific tax breaks, and politicians lobby for keeping certain territorialTE schemes, such as special economic zones (SEZs), that favour their electoraldistricts or local constituencies (for instance, see Daude et al. 2017).

The Inter-American Center of Tax Administrations (CIAT) Tax ExpenditureDatabase (TEDLAC) shows that, on average, TE as percentage of GDP amountedto 3.5% in the region, ranging from 0.7% in Colombia to 6.6% in DominicanRepublic (Pelàez Longinotti 2018). Kassim and Mansour (2018) review TEreporting in 26 low- and middle-income economies and show that the revenueforegone through these provisions is in most of the cases significant. In their sample,TE as a share of GDP ranges from 1.38% in Burkina Faso to 4.69% in Poland and6.13% in Ghana. Likewise, TE as a share of total tax revenue is as high as 27.8% inPoland, 41.67% in Ghana and a staggering 58% in Mauritania.

It has been argued that, despite their fiscal cost, TEs can be a valid policyinstrument for governments worldwide. Indeed, under certain conditions(e.g. when eligibility conditions are directly linked to tax return data, when it ismore important to maximise the number of beneficiaries than to minimize excessclaims or when the policy objective is to incentivize a clear and broadly definedactivity by reducing its net price), TEs could be more cost-effective than directspending and may hence be the best option to pursue a specific public policy goal(Toder 2000).

176 A. Redonda et al.

Yet, TE provisions are generally ineffective in reaching their stated goals. The useof tax incentives to attract investments is a case in point. Low- and middle-incomeeconomies often grant far-reaching tax holidays and tax exemptions that have littleimpact on investment or growth and, in addition, significantly reduce the availabilityof public funds (IMF et al. 2015).

To sum up, scaling back ineffective TEs would broaden tax bases and, at the sametime, allow countries to reduce marginal tax rates and/or save resources that, in turn,could be allocated to more productive or more effectively targeted purposes. How-ever, a necessary condition for a sound evaluation of the cost and impact of theseprovisions is data (in particular, revenue foregone estimates), which—when exis-tent—are generally published in governmental TE reports or statements. While it iscertainly true that public reporting alone is not sufficient to achieve more rationalityin TE, it is hard to engage in meaningful discussion on individual TE provisionswithout access to reliable, detailed and timely data.

9.3 Tax Expenditure Reporting

Systematically identifying, estimating and reporting the fiscal cost of TEs can be atime- and resource-intensive task. As discussed by Heady and Mansour (2019),countries could adopt a gradual approach that takes their institutional and datacapacities into account. The returns to such a process are potentially high, since itis critical to assess the effectiveness and efficiency of these provisions and, hence, toensure the alignment of public policies with their stated goals. In addition, publish-ing comprehensive and detailed TE reports is crucial not only as a policy tool forpolicymakers but also as a means for increasing government transparency andaccountability.

Yet, despite their significant impact on government budgets, TEs are generallymuch less an object of public scrutiny than direct spending. As acknowledged by theAustralian Treasury, “Tax Expenditures, like direct expenditures, affect the govern-ment’s budget. However, unlike direct expenditures, tax expenditures once legis-lated become part of the tax law with a recurring fiscal impact and do not receiveregular scrutiny through the budget process” (Australian Treasury 2004). Indeed, TEreporting lags way behind best practice in most countries, and several countries thatissue reports do not even estimate the revenue foregone through these provisions. Anassessment of TE reporting in the 43 Group of 20 (G20) and OECD economies,based on 9 key dimensions that reflect good practice in TE reporting, shows that8 countries have not reported on TEs in the last 10 years, 26 have published a basicreport (e.g. by providing estimates for a reduced subset of TEs or estimates based onaggregate figures only) during the same period and only 9 governments havepublished a detailed and comprehensive TE report on a regular basis (Redondaand Neubig 2018).

Furthermore, the same assessment shows that among the 43 G20 and OECDeconomies, TE under-reporting (i.e. difference between the number of effectively

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 177

estimated TEs and the number of those provisions that are listed without an estimateof their fiscal cost) is highly prevalent. For instance, the share of estimated TEs withrespect to listed provisions is as low as 17% in Greece and 20% in New Zealand(Redonda and Neubig 2018). Yet, a larger ratio of estimated/listed provisions doesnot necessarily indicate that the report is more comprehensive. As will be illustratedwith the South African example later in the chapter, the number of TEs included inthe reports (estimated and/or listed) is very often far from reflecting the number ofTEs effectively implemented by the government.

Moreover, while there is significant room for improving the estimation andreporting of TEs in advanced economies, the situation is even more worrisome inmiddle- and low-income countries. Indeed, in most of these countries, TE reportingis rudimentary at best, due to data constraints, insufficient human and financialresources as well as weak institutions, among other factors. According to figuresprovided by the World Bank in 2015, the percentage of countries that periodicallyestimates the revenue forgone through TEs amounts to 33% in Latin America andthe Caribbean, 21% in sub-Saharan Africa, 14% in South Asia and 10% or less in theMiddle East and North Africa, Europe and Central Asia as well as East Asia and thePacific. Likewise, Kassim and Mansour (2018) show that TE reporting qualityremains considerably weak in several areas in most of the 26 low- and middle-income countries covered by their study. For instance, only about half of thereviewed countries have a legal requirement to report on TEs, only three reportsprovide forecast estimates for future years and roughly half of the countries provide adiscussion of the benchmark system in their reports.3

9.4 Tax Expenditure Reporting in Africa

As shown in Fig. 9.1, out of 54 African countries reviewed by the Global TaxExpenditures Database (GTED) team, 20 (about 37% of the countries) reported theirTE to the public at least once between the years 2000 and 2019.4 The remaining 63%of the countries did not publish any TE reports during this period. We start our

3This information is crucial since TEs are defined as deviations from a benchmark, which representsthe standard taxation treatment and, at the same time, could include certain elements of the taxsystem which depart from a uniform treatment of taxpayers, i.e. fundamental structural elements ofthe tax system such as the progressive income tax rate scale for individual taxpayers (AustralianTreasury 2011).4The GTED is a joint effort of think tanks and research institutions from Europe, Asia, Africa andLatin America, led by the Council on Economic Policies (CEP) and the German DevelopmentInstitute (DIE) and financed by the German Federal Ministry for Economic Cooperation andDevelopment (BMZ). The main goal of the project is to build the first online Global TaxExpenditures Database to increase transparency, generate trustworthy information and expandresearch in the field of TEs. The GTED is set up with official data on TEs published by governmentsworldwide, in a consistent format to increase the level of international comparability. The GTEDwill be made publicly available and free of charge through an online platform that will also include a

178 A. Redonda et al.

discussion by taking a more detailed look at this second group before zooming in onthe 20 countries that provided at least some information.

9.4.1 African Countries Without Publicly Available TEReports

For the purposes of transparency and political debate, TE reports have to be public.Ideally, they should be found on open and easily accessible governmental websitesor repositories. References to such reports in official government statements shouldinclude information on where to find them. For 6 out of the 34 countries classified asnot having a public TE report, we found references to internal TE reports, but nofurther information was given. One such example is Ghana. In the FY 2019 BudgetSpeech, the Ghanaian Finance Minister states that “[between 2010 and 2018], aboutsix different studies have been conducted into Ghana’s tax exemptions regime by theRevenue Policy Division of the Ministry of Finance, International Monetary Fund

15

4

6

11

17

0

5

10

15

20

25

30

35

40

Repor�ng Not Repor�ng

TE Repor�ng in Africa: Number of Countries

Itemized TE Report Only Overall TE Es�mates TE Report Not Public

TE Report Underway No TE Informa�on

Fig. 9.1 Overview of TE reporting in Africa (“Itemized TE report” refers to TE reports that providesubcategorization beyond tax base information. The most advanced version of an itemized TEreport is one that provides revenue foregone and accompanying information at the TE provisionlevel.). Source: Global Tax Expenditures Database, www.GTED.net

repository for blogs, working papers and further publications as well as events on TEs organized bythe partners. For more details, see Redonda and von Haldenwang (2021).

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 179

(IMF), Organisation for Economic Cooperation & Development (OECD), GIZ andthe World Bank” (Ministry of Finance, Republic of Ghana 2018). However, we werenot able to find any of these reports. Similarly, the Eswatini (formerly known asSwaziland) Revenue Authority claims that it carries out a TE analysis biennially.“[Exemptions] under Income Taxes are reported every 2 years after carrying out aTax Expenditure Analysis. The last report done was for 2013/14 and the next reportwill be undertaken in the new financial year, 2016/17” (Eswatini Revenue AuthorityAnnual Report 2016). However, these reports are also not available.

Furthermore, 11 countries claim to be in the process of carrying out a TE analysisor are considering the possibility of doing so. Chad, for example, is reported to havestarted a TE analysis in 2016 but has since not published any information regardingthat particular analysis or TE in general. “Since September 2016, with the support ofthe World Bank EFSO [Emergency Fiscal Stabilization Development Policy Oper-ation], the [Chad] Government put a freeze on any new tax exemptions, except thoseunder the Vienna convention or those related to PTF [Partnership TransparencyFund] until an ongoing study on tax expenditure would be completed and newmeasures adopted” (World Bank 2018). Similarly, Ethiopia is also reported to have aTE analysis “near completion,” but additional information about the analysis is notavailable (IMF 2018). Table 9.1 provides information on the TE reporting state ofthe art in each of the assessed countries.

Lastly, the remaining 17 countries do not have any information regarding TEs onany online platform. In the case of the Central African Republic, for example, theIMF reports, “[tax] exemptions and tax expenditures are rarely detailed and assessedin the budget documentation or are maintained for several years without clear legalbasis and economic justification” (IMF 2019a). Similarly, in Kenya, “[there] is noreporting of tax expenditures. Despite a constitutional requirement for a publicrecord of tax waivers to be published, and the PFM Act’s requirement for an annualreport, no such reports are available” (IMF 2017b). This shows that some govern-ments even ignore legal or constitutional norms on TE analysis and reporting.

9.4.2 African Countries with Publicly Available TE Reports

A total of 20 African countries reported on their TEs at least once in the last 20 years.However, the information reported during this period varies substantially acrosscountries. Countries like the Democratic Republic of the (DR) Congo, Lesotho,Seychelles, Sierra Leone and South Sudan only provide very rudimentary estimatesof TEs in budget-related documentation, without any further details. For example, inthe FY 2019 Budget Speech, the South Sudan Finance Minister claimed that“exemptions have reached recorded level of sixty-nine billion South SudanesePounds, which deprived the government of the most resources for capital investmentand development” (Ministry of Finance and Planning, Republic of South Sudan2018). However, the ministry did not provide any information on how it calculatedthose estimates. This is a case that could also qualify as non-reporting. Similarly, the

180 A. Redonda et al.

Lesotho Revenue Authority Annual report includes a small section with an overallestimate on “Revenue Foregone on Tax Relief Measures,” or similarly titled (Leso-tho Revenue Authority Annual Report 2016). However, the report does not breakdown the overall TE estimate or provide a clear description on how this estimate wascalculated. The General Directorate for Imports of DR Congo also publishes overallTE estimates in several of its annual reports with little accompanying informa-tion (Ministère des Finances et du Budget, Republique du Congo 2017).

In fact, only 15 African countries have ever published itemized reports. None-theless, as shown in Fig. 9.2, this number is steadily rising. The first TE report inAfrica was published in Morocco in 2005. Two years later, Mauritius also startedreporting on TEs. South Africa started reporting only in 2011. In 2018, the numberof reports published in Africa reached a record of seven reports. Nonetheless, thenumbers also highlight the inconsistency of TE reporting among African countries.As shown in Fig. 9.2, 10 out of the 15 African countries classified as having reportedon TEs had only ever published one or two reports. Two countries (Mali andSenegal) published between three and five reports, and only three governments—Mauritius, Morocco and South Africa—publish TE reports regularly. Tanzania is aspecial case since it has covered 12 years of TE estimates in one single TE report.

Table 9.1 Overview of TE reporting in Africa, by country

TE report available

TE report notpublic

TE report not available

Itemized TEreport

Only overall TEestimates

TE reportunderway No TE information

Benin DR Congo Cameroon Algeria Angola

Burkina Faso Lesotho Ghana Botswana Burundi

Ivory Coast Seychelles Mozambique Chad Cape Verde

Gabon Sierra Leone Swaziland Egypt Central AfricanRepublic

Guinea South Sudan Uganda Ethiopia Comoros

Liberia EquatorialGuinea

Gambia Congo

Madagascar Guinea-Bissau Djibouti

Mali Niger Eritrea

Mauritania Nigeria Kenya

Mauritius Togo Libya

Morocco Zambia Malawi

Rwanda Namibia

Senegal Sao Tome andPrincipe

South Africa Somalia

Tanzania Sudan

Tunisia

Zimbabwe

Source: Global Tax Expenditures Database, www.GTED.net

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 181

However, even for these reporting countries, much progress is needed in terms ofcoverage.

The quality and the scope of most TE reports do not match the standardsdeveloped, for instance, by the IMF (2019b). Morocco and Ivory Coast are notableexceptions due to the broad scope of information provided in their reports. Moroccopublishes in-depth information—detailed description, legal reference, tax base,incentive mechanism, budgetary category, targeted beneficiaries, policy objectiveand revenue foregone—for each of its TE provisions. Ivory Coast publishes the samekind of detailed information but does only include aggregate estimates of revenueforegone. Both countries provide much of the information that can be found in themost advanced and in-depth reports of countries such as Australia and Canada. Onlyinformation regarding the number of beneficiaries of each TE provision is missingfrom the TE reports of these two countries.

Apart from these two cases, most of the countries with TE reports providedrevenue foregone estimates that were aggregated either at the tax base or at thebudgetary category level, without supplying specific information at the TE provisionlevel. Such information would be required, however, to perform cost-benefit eval-uations and hence assess the effectiveness and efficiency of these provisions.Moreover, no country presents consistent data on the number of beneficiaries perprovision.5 Less than half of the countries specified the budgetary category, revealeda policy objective, specified the targeted beneficiaries or included a detailed descrip-tion for each TE provision. This kind of information is not only crucial forpolicymakers, but it is also important to increase transparency and accountability

0

2

4

6

8

10

12

1 to 2 3 to 5 More than 5

Num

ber o

f Cou

ntrie

s

Number of Reports

A

0

1

2

3

4

5

6

7

8

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

Num

ber o

f Rep

orts

Year

B

Fig. 9.2 Number of reports per country and per year. Source: Global Tax Expenditures Database,www.GTED.net

5Rwanda provides information on the number of beneficiaries for some provisions, but not for most.Hence, it received a negative score in the evaluation.

182 A. Redonda et al.

towards the society. On the other hand, most countries did present information onmethodology, the legal reference and the tax base of specific TE provisions.6

Following Kassim and Manosur (2018) and Redonda and Neubig (2018), weevaluate the TE reports of the 15 reporting countries. As shown in Table 9.2,Morocco, Burkina Faso and Ivory Coast provide the most detailed TE informationat the provision level. On the other hand, Mauritius and South Africa are thecountries providing the least detailed information at the provision level. However,the difference between the South African report and the Mauritian report is that theformer publishes some information at the TE provision level and describes itsmethodology, whereas the latter only publishes an itemized table regarding TE inits budget-related documentation without any accompanying information.

Table 9.2 highlights the heterogeneity and, at the same time, the overall need forimprovement in African TE reporting. For example, countries such as Mauritius,South Africa, Gabon and Liberia did not provide any information at the provisionlevel. The Gabon report barely fits the definition of an “itemized report” as it onlylists nine TE subcategories and only reports on TE from value-added tax (VAT).Moreover, countries such as Benin, Guinea, Madagascar and Senegal providerevenue foregone estimations at the provision level but do not provide any othercompanion information such as a policy objective. These two pieces of informationare the most important when conducting cost/benefit analyses yet rarely reportedtogether by African countries. Indeed, only 3 out of 15 African TE reports presentboth revenue foregone estimates and policy objectives at the TE level.

Finally, even though we observe an increase in the number of countries reportingon TEs, the overall quality of the reports has not improved across years. Reportspublished prior to 2018 and those published in 2018 or 2019 received roughly thesame score in the evaluation. Mauritius, Morocco and South Africa are the onlycountries for which an individual evaluation across time is possible. The threecountries show barely any change in the quality and scope of their reporting,i.e. their first and last TE reports followed the same format. While this may be anindicator of consistency, it is also obvious that a large amount of relevant informa-tion is still missing in most reports.

9.4.3 What the Current Reports Reveal About TE in Africa

Some key messages can be drawn from the current TE reports from Africa. Table 9.3shows each reporting country’s total TE as a percentage of the total amount of taxescollected and as a percentage of each country’s GDP during the same year. Onaverage, the revenue foregone on TE among African countries amounted to 17.8% of

6Most countries did not provide methodology information at the provision level. However, mostgovernments that do report use only one methodology to calculate the revenue foregone for all TEprovisions, which is then described in one section of the report.

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 183

Tab

le9.2

Evaluationof

TErepo

rting,

latestavailableyear

Cou

ntry

Revenue

forego

neMetho

dology

Detailed

description

Legal

reference

Tax

base

info

Incentive

mechanism

Bud

getary

catego

ryTargeted

beneficiaries

Num

berof

beneficiaries

Policy

objective

Benin

(201

9)✓

✓✓

✓✓

Burkina

Faso

(201

6)✓

✓✓

✓✓

✓✓

✓

IvoryCoast

(201

7)✓

✓✓

✓✓

✓✓

✓

Gabon

(201

8)✓

Guinea

(201

7)✓

✓✓

✓✓

Liberia

(201

6)

Madagascar

(201

5)✓

✓✓

✓

Mali(201

9)✓

Mauritania

(201

3)✓

✓✓

✓

Mauritiu

s(201

6)

Morocco

(201

9)✓

✓✓

✓✓

✓✓

✓✓

Rwanda

(201

8)✓

✓✓

✓✓

✓

Senegal

(201

4)✓

✓✓

✓✓

✓✓

Sou

thAfrica

(201

8)✓

184 A. Redonda et al.

Tanzania

(201

8)✓

✓✓

✓✓

Sou

rce:GlobalTax

Exp

endituresDatabase,www.GTED.net

The

evaluatio

nisbasedon

themostrecent

repo

rtof

each

coun

try—

year

inparenthesis

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 185

total tax collected. However, the TE share of total tax collected varied substantiallyacross countries. In countries like Mauritania and Senegal, for example, this sharewas as high as 58.4% or 39.6%, respectively. Meanwhile, in Burkina Faso, Sey-chelles and Sierra Leone, TE was less than 9% of total tax collected. Similarly, TE asa share of GDP also varied across African countries. On average, TE accounted for2.7% of GDP across African countries. However, this share was as low as 0.3% inthe Seychelles and as high as 7.8% in Senegal. Furthermore, it is important to notethat some countries such as Gabon, Liberia and Sierra Leone only reported their TEfor one specific tax base (VAT in the case of Gabon and customs for the rest). Hence,the share of TE in total tax collected or in total GDP for these countries is likelyhigher.

As shown in Table 9.4, TE across African countries came from similar tax bases.VAT expenditure comprised the largest share of TE in most African countries. Onaverage, almost half (49%) of TE across African countries originated fromVAT-related exemptions and rate reductions. Customs duties were also a big sourcefor TE across Africa. More than one-fifth (20% on average) of TE stemmed from

Table 9.3 TE as a percentageof total tax collected and GDP

Country Year % of tax collected % of GDP

Benin 2017 18.20 2.40

Burkina Faso 2016 8.61 1.38

DR Congo 2016 10.21 0.65

Ivory Coast 2017 9.80 1.32

Gabon 2017 12.10 1.24

Guinea 2017 21.70 2.63

Lesotho 2016 17.60 3.96

Liberia 2016 18.89 4.70

Madagascar 2015 17.00 1.79

Mali 2017 17.32 2.64

Mauritania 2013 58.41

Mauritius 2017 9.22 1.76

Morocco 2018 13.01 2.78

Rwanda 2018 14.30 4.60

Senegal 2014 39.60 7.80

Seychelles 2019 1.04 0.34

Sierra Leone 2017 8.76 1.20

South Africa 2017 14.90 3.90

Tanzania 2012 27.00 4.40

Average 17.77 2.75

Source: Global Tax Expenditures Database, www.GTED.netIncludes the countries which provided only overall estimatesSouth Sudan only provided an overall estimate of total “exemp-tions” in local currency. However, due to a drop of GDP by 80%during the same reporting year, it is hard to calculate a meaningfulTE/GDP ratio. Similarly, due to a steep currency devaluation inMauritania between 2010 and 2017, it is hard to calculate ameaningful TE/GDP ratio for that country as well

186 A. Redonda et al.

customs-related taxes.7 Income tax-related expenditure was also important, eventhough its significance varied by country. For example, personal income tax (PIT)expenditure was less than 1% of TE in Benin and Madagascar, while it was morethan 35% in South Africa and as high as 71.2% in the Seychelles. Similarly,corporate income tax (CIT) expenditure was lower than 3% in Benin and Madagas-car but higher than 45% in Mauritius. However, the presentation of income TEvaried by country. While 11 countries reported CIT and PIT separately, othercountries did not differentiate between CIT and PIT. In these cases, income tax TEwas included in the “other” category. Moreover, revenue foregone from excise taxesand especially stamp duties usually comprised a small portion of TE (except forstamp duty TE in the case of Benin). Lastly, taxes such as apprenticeship taxes,export taxes and other country-specific taxes also comprised a small portion of TEand were included in the “other” category.

In addition, while beneficiary information was absent from a lot of African TEreports, some insight can nonetheless be drawn from the available information. Asshown in Table 9.5, businesses benefited the most from African TE. On average, a

Table 9.4 Share of total revenue foregone by tax base

Country YearVAT(%)

PIT(%)

CIT(%)

Customs(%)

Stamp duty(%)

Excise(%)

Others(%)

Benin 2017 62.7 0.1 1.5 10.3 25.0 0.1 0.3

BurkinaFaso

2016 40.7 21.0 26.1 8.5 0.0 3.8

DR Congo 2016 66.7 33.3

Ivory Coast 2017 69.6 9.2 0.7 20.5

Gabon 2017 100.0 0.0

Guinea 2017 40.0 34.0 11.0 15.0

Lesotho 2016 38.2 61.8 0.0

Liberia 2016 100.0 0.0

Madagascar 2015 24.2 0.5 2.5 42.0 8.0 22.7

Mali 2017 60.3 11.7 24.2 2.7 1.1

Mauritania 2013 49.1 9.6 4.4 27.1 0.7 9.1

Mauritius 2017 33.5 8.0 46.0 2.3 10.2 0.0

Morocco 2018 52.9 15.9 15.6 2.7 2.4 0.9 9.5

Rwanda 2018 60.7 27.5 11.8

Senegal 2014 45.8 22.5 11.4 9.7 2.8 7.8

Seychelles 2019 4.6 71.2 18.5 5.7 0.0

Sierra Leone 2017 100.0 0.0

South Africa 2017 35.4 36.6 4.7 23.3 0.0

Tanzania 2012 44.0 56.0

Source: Global Tax Expenditures Database, www.GTED.net

7The numbers from Gabon, Liberia and Sierra Leone were removed when calculating thoseaverages since these countries only reported their TE for one singular tax base.

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 187

little over 50% of TE across African countries stemmed from exemptions and otherincentives given to businesses. Households were also important beneficiaries ofTE. They received a little over 25% of the TE. However, the TE share of businessesand households in some countries is even higher since they reported the TE given tothose two beneficiaries jointly, on top of reporting their individual shares (in suchcases, the joint share is added to the “other” category in Table 9.5). Moreover, whilethe government was often an important beneficiary of TE, intergovernmental orga-nizations (IGOs) and non-governmental organizations (NGOs) were usually not.Lastly, country-specific beneficiaries such as the mining sector or TE provisionswith missing beneficiary information were categorized as “other.”

Yet, it is important to note that the numbers reported in Tables 9.3, 9.4 and 9.5must be interpreted with caution. As shown in Fig. 9.3, countries do not estimate therevenue foregone for all TE provisions they report. Hence, the numbers reportedabove either understate or overstate the significance of some categories and the totalmagnitude of TE. For example, while two-fifths of the sample countries do notspecify the share of TE provisions evaluated, some countries such as Benin andMadagascar report that they evaluated less than 50% of their TE provisions. BurkinaFaso and Mauritania report that they evaluated around 60% of their TE provisions.Lastly, Ivory Coast, Guinea, Morocco and Senegal report having evaluated 70% ormore of the TE provisions identified in their tax codes. However, even if thecountries claim to have evaluated the revenue foregone from a majority of theirprovisions and their TE/GDP ratio may seem realistic, often, a portion of their TEprovisions is not included in their reports.

Table 9.5 Share of revenue foregone by beneficiary type

Country YearPublicsector (%)

Businesses(%)

Households(%)

NGOs(%)

IGOs(%)

Others(%)

Benin 2017 23.1 39.6 34.0 1.6 1.7

BurkinaFaso

2016 10.1 75.5 11.3 0.0 3.1

Ivory Coast 2017 28.6 63.5 0.8 4.3 2.8

Gabon 2017

Guinea 2017 64.8 2.6 32.6

Liberia 2016

Madagascar 2015 79.6 0.6 19.9

Mali 2017

Mauritania 2013 61.4 38.6

Mauritius 2017

Morocco 2018 2.7 47.4 48.4 1.5

Rwanda 2018

Senegal 2014 5.2 20.5 61.7 7.2 5.4

South Africa 2017

Tanzania 2012 17.1 7.8 0.4 74.7

Source: Global Tax Expenditures Database, www.GTED.net

188 A. Redonda et al.

South Africa is a case in point. The latest TE statement published by the treasuryreports estimates for 34 provisions, 13 TEs granted through PIT, 9 through CIT,7 through VAT and 5 through customs and excise taxes. Total TEs in 2016, the latestavailable year, were estimated at Rands 209,007 accounting for 18.3% of total grosstax revenue and 4.7% of GDP.

Yet, these figures should be interpreted cautiously since the statement does notinclude several TEs. For instance, individuals working overseas for a 183-day termcan claim back tax deductions on income earned for the period they were outside thecountry. In addition, as PTEs, tax-free savings accounts (TFSAs) aim at boostingindividuals’ savings by fully exempting the money invested into a TFSA from anytax on interests, dividends, capital gains and withdrawals. Moreover, individuals candeduct various work-related expenses. For example, individuals using their vehiclefor work purposes can also claim a deduction. The depreciation on devices(e.g. laptops, smart phones) bought and maintained in one’s personal capacity forwork may also be claimed as a tax deduction. Finally, employers that work fromhome on a regular basis and in a specifically dedicated space are also allowed toclaim certain running costs associated to that space, including rent, electricity andmaintenance expenses.

When it comes to businesses, income from international shipping of a residentcompany that holds a share in a South African-flagged ship is exempt from CIT.Moreover, the Headquarter Company (HQC) Regime eliminates multiple fiscal andregulatory barriers to encourage foreign investors to use South Africa as a holdingcompany/HQC jurisdiction. Among other benefits, this regime exempts eligiblebusinesses from dividends tax and capital gains taxes. Finally, as in several low-and middle-income economies, South Africa created several SEZs, i.e. designedareas for targeted economic activities where a myriad of tax benefits are in place.These benefits include a reduced 15% CIT rate (compared to the standard 28% rate),

0%10%20%30%40%50%60%70%80%90%100%

0

100

200

300

400

500

600

% o

f Tot

al L

isted

# of

TE

Prov

ision

s

Country

Listed Es�mated Share of es�mated TEs

Fig. 9.3 Percentage of TE provisions evaluated. Source: Global Tax Expenditures Database, www.GTED.net

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 189

a 10% building allowance, an employment tax incentive as well as additional VATand duties reliefs for firms operating within a customs-controlled area of an SEZ.8

The list of existing TEs that are not covered by the TE statement is likely to bemuch broader for other countries as well. A first step towards a more transparentreporting on TEs would be to include all existing provisions even when, for some ofthem, revenue foregone estimates are not provided. Ideally, the report will progres-sively increase the number of estimated provisions in order to increase transparencyand accountability.

9.5 Conclusion

This chapter has discussed the role of TE reporting as a key component of a moreeffective and fairer tax system. In addition, it has given an overview of TE reportingin Africa, based on data provided by the GTED. With few exceptions, TE reportingis at a very initial stage in most countries of the region. As a result, citizens andpolitical decision-makers in a majority of African countries do not have access totimely and detailed information on a set of highly relevant policy instruments thatcause major drains to the public budget and, consequently, cannot assess the costsand benefits that these provisions generate for society. With anecdotal evidencedominating the public debate, under-reporting of existing TE regimes tends tofurther undermine the legitimacy and perceived fairness of taxation.

While Africa is the only region fully covered by the GTED at this moment, it isnot at all clear that the other world regions perform significantly better. On the onehand, TE reporting is much more widespread in Latin America, as data presented byCIAT shows.9 On the other hand, other regions—Asia in particular—are ratheruncharted territory in terms of TE reporting as well. Even among the OECD andG20 countries, the frequency and quality of reports are highly heterogeneous, asRedonda and Neubig (2018) observe.

It should be noted, however, that the TE topic has recently attracted moreattention by scholars, international organisations and civil society. Efforts such asthe Latin America Tax Expenditure Research, Advocacy, and Learning (LATERAL)project led by the International Budget Partnership (IBP)10 aim at covering existingknowledge gaps and raising public awareness of the issue. Several internationalorganisations have produced reports and guidelines that aim at improving TEreporting (for instance, see IMF 2019b). Yet, much more remains to be done.What is still missing at this stage is decisive action to agree on joint internationalstandards for the robust, public, exhaustive and timely assessment of the costs and

8Although the total fiscal cost of SEZs is difficult to be estimated, some of these provisions such asthe Employment Tax Incentive (ETI) may be included in the TE Statement.9See https://www.ciat.org/tax-expenditures/?lang¼en, accessed 13.09.2019.10See https://www.internationalbudget.org/analysis-insights/lateral-project/, accessed 18.09.2019.

190 A. Redonda et al.

benefits of TEs. We believe that the GTED will make a substantial contributiontowards this end.

References

Alstadsæter, A., Barrios, S., Nicodeme, G., Skonieczna, A., & Vezzani, A. (2018). Patent boxesdesign, patents location, and local R&D. Economic Policy, 33(93), 131–177.

Australian Treasury. (2004). Tax expenditures statement 2003. https://treasury.gov.au/publication/tax-expenditures-statement-2003

Australian Treasury. (2011). Tax expenditures statement 2010. https://treasury.gov.au/publication/tax-expenditures-statement-2010/tax-expenditures-statement-2010

Australian Treasury. (2018). Tax expenditures statement 2017.https://treasury.gov.au/sites/default/files/2019-03/2017-TES.pdf

Axelson, C., Pomeranz, D., & Redonda, A. (2020). The distributive impact of pension-related taxexpenditures in South Africa. mimeo.

Bird, R. M., & Zolt, E. M. (2015). Fiscal contracting in Latin America. World Development, 67,323–335.

Brys, B., Perret, S., Thomas, A., & O’Reilly, P. (2016). Tax design for inclusive economic growth.OECD Taxation Working Paper 26. OECD Publishing.

Chelminski, K. (2018). Fossil fuel subsidy reform in Indonesia. In J. Skovgaard & H. van Asselt(Eds.), The politics of fossil fuel subsidies and their reform (pp. 193–211). Cambridge: Cam-bridge University Press.

Daude, C., Gutierrez, H., & Melguizo, A. (2017). Doctoring the ball: The political economy of taxincentives for investment in the Dominican Republic. Journal of Economic Studies, 44(1), 2–23.https://doi.org/10.1108/JES-05-2015-0090.

Département Fédéral des Finances. (2011). Quels Sont les Allégements Fiscaux Accordés par laConfédération? Département Fédéral des Finances (DFF) de la Confédération Suisse(in French). https://www.estv.admin.ch/dam/estv/fr/dokumente/allgemein/Dokumentation/Zahlen_fakten/berichte/2011/Welche%20Steuerverg%C3%BCnstigungen%20gibt%20es%20beim%20Bund%20Bericht.pdf.download.pdf/2011_Steuerverguenstigung_f.pdf

Direction du Budget. (2019). Évaluations des Voies et Moyens. Tomme II. Dépenses Fiscales.Annexe au Projet de Loi des Finances pour 2019 (in French). https://www.performance-publique.budget.gouv.fr/sites/performance_publique/files/farandole/ressources/2019/pap/pdf/VMT2-2019.pdf

Duflo, E., Gale, W., Liebman, J., Orszag, P., & Saez, E. (2006). Saving incentives for low- andmiddle-income families: Evidence from a field experiment with H&R block. The QuarterlyJournal of Economics, 121(4), 1311–1346.

Eswatini Revenue Authority. (2017). Eswatini revenue authority annual report, 2016. http://www.sra.org.sz/documents/ANNUAL%20REPORT%202016.pdf

Heady, C., & Mansour, M. (2019). Tax expenditures and their use in fiscal management: A guide todeveloping countries. How to Note 19/01, International Monetary Fund (IMF). https://www.imf.org/~/media/Files/Publications/HowToNotes/HTNEA2019002.ashx

Hilber, C., & Turner, T. (2014). The mortgage interest deduction and its impact on homeownershipdecisions. Review of Economics and Statistics, 96(4), 618–637.

IMF. (2017a, April). Fiscal monitor. Achieving more with less. IMF. http://www.imf.org/en/Publications/FM/Issues/2017/04/06/fiscal-monitor-april-2017

IMF. (2017b). Kenya fiscal transparency evaluation. International Monetary Fund. https://www.imf.org/external/pubs/ft/scr/2016/cr16221.pdf

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 191

IMF. (2018). The federal Democratic Republic of Ethiopia staff report for the 2017 article IVconsultation. International Monetary Fund. https://www.imf.org/~/media/Files/Publications/CR/2018/cr1818.ashx

IMF. (2019a). Central African Economic and Monetary Community (CEMAC): Selected issues. Aregional approach to enhancing governance and reducing the potential for corruption. Inter-national Monetary Fund. https://www.elibrary.imf.org/view/IMF002/25705-9781484392843/25705-9781484392843/25705-9781484392843_A001.xml?redirect=true

IMF. (2019b). Tax expenditure reporting and its use in fiscal management. A Guide for DevelopingEconomies How To Notes 19/01. Washington, DC: International Monetary Fund.

IMF, OECD, UN, & World Bank. (2015). Options for low income countries’ effective and efficientuse of tax incentives for investment. A Report to the G-20 Development Working Group. https://www.imf.org/external/np/g20/pdf/101515.pdf

Kassim, L., & Mansour, M. (2018). Les Rapports sur les Dépenses Fiscales des Pays enDéveloppement : Une Evaluation. Revue d’économie du développement, 26(2), 113–167.(in French).

Klemens, B. (2017). Intellectual property boxes and the paradox of price discrimination. CEPWorking Papers 1703, Council on Economic Policies (CEP).

Lesotho Revenue Authority. (2017). Lesotho revenue authority annual report 2016. http://passthrough.fw-notify.net/download/860771/http://www.lra.org.ls/sites/default/files/2018-02/LRA%20Annual%20Report%20-%202016.pdf

Ministère des Finances et du Budget, Republique du Congo. (2017). Loi de Finances 2018(French). https://www.finances.gouv.cg/fr/loi-de-finances-2018

Ministry of Finance and Planning, Republic of South Sudan. (2018). Budget speech FY 2018/19.http://grss-mof.org/wp-content/uploads/2018/09/Budget-Speech-Final-July-12-_-12-05-pm.pdf

Ministry of Finance, Republic of Ghana. (2018). The budget statement and economic policy of thegovernment of Ghana for the 2019 financial year. http://www.mofep.gov.gh/sites/default/files/budget-statements/2019-Budget-Statement-and-Economic-Policy_.pdf

Pelàez Longinotti, F. (2018). Overview of tax expenditures in Latin America: Main statistics of theCIAT database. Working Paper, Inter-American Center of Tax Administrations (CIAT).

Prichard, W. (2019). Tax, politics, and the social contract in Africa. Oxford: Oxford UniversityPress.

Redonda, A. (2016). Tax expenditures and sustainability. An overview. CEP Discussion Note16/03, Council on Economic Policies (CEP).

Redonda, A., & von Haldenwang, C. (2021). CEP/DIE Companion to the Global Tax ExpendituresDatabase, GTED WP 2021/1. mimeo.

Redonda, A., & Neubig, T. (2018). Assessing tax expenditure reporting in G20 and OECDeconomies. CEP Discussion Note 2018/3, https://www.cepweb.org/assessing-tax-expenditure-reporting-in-g20-and-oecd-economies/

South African National Treasury. (2012). Improving tax incentives for retirement savings. Techni-cal Discussion Paper E for Public Comment. http://www.treasury.gov.za/comm_media/press/2012/Improving%20tax%20incentives%20for%20retirement%20savings.pdf

Toder, E. (2000). Tax cuts or spending—Does it make a difference? National Tax Journal, 53,361–371.

UNCTAD. (2014). World investment report 2014. Investing in the SDGs: An action plan. UnitedNation Conference on Trade and Development (UNCTAD). https://unctad.org/en/PublicationsLibrary/wir2014_en.pdf

US Treasury. (2020). Tax expenditures. https://home.treasury.gov/policy-issues/tax-policy/tax-expenditures

von Haldenwang, C. , & Laudage, S. (2019). Financing for development and domestic revenuemobilisation: More international reforms are needed. Briefing Paper 13/2019. Bonn: DIE.

World Bank. (2015). World Bank East Asia and Pacific Economic Update. Staying the course.World Bank (WB). http://pubdocs.worldbank.org/en/414911444005973491/pdf/EAP-Economic-Update-2015-10.pdf

192 A. Redonda et al.

World Bank. (2017). Special economic zones in the Dominican Republic: Policy considerations fora more competitive and inclusive sector: Special economic zones, global value chains, and thedegree of domestic linkages in the Dominican Republic (Spanish), http://documents.worldbank.org/curated/en/734421487332577036/Special-economic-zones-global-value-chains-and-the-degree-of-domestic-linkages-in-the-Dominican-Republic

World Bank. (2018). Republic of chad, first economic recovery and resilience development policyoperation. World Bank. http://documents.worldbank.org/curated/en/416901537280719716/pdf/116242-PGD-OUO-9-Chad-1st-Economic-Recovery-Resilience-DPO-PD-August-14.pdf

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,adaptation, distribution and reproduction in any medium or format, as long as you give appropriatecredit to the original author(s) and the source, provide a link to the Creative Commons license andindicate if changes were made.

The images or other third party material in this chapter are included in the chapter’s CreativeCommons license, unless indicated otherwise in a credit line to the material. If material is notincluded in the chapter’s Creative Commons license and your intended use is not permitted bystatutory regulation or exceeds the permitted use, you will need to obtain permission directly fromthe copyright holder.

9 Tax Expenditure Reporting and Domestic Revenue Mobilization in Africa 193